Issue link: https://iconnect007.uberflip.com/i/1327102
JANUARY 2021 I DESIGN007 MAGAZINE 21 Another common problem we see is on heavier copper stackups. In Figure 7 we can see the heavier copper gets taller, and we need to space out these signals, which are basically lines running in parallel to each other, in order to etch them. With thicker copper, we need wider lines and spaces. Often, when custom- ers have problems with their designs, it's with an established design. They're trying to adapt it by increasing the copper thicknesses on the layers, and the lines and spaces were already at the maximum that they could be that fit in the circuitry, but then we can't etch them. An issue that makes stackups so difficult is the predictability. When we put down these cores, and we melt the prepregs between them under high temperature and pressure, they might start out at a certain thickness, but how thick are they going to mush together after lamination? For heavier Cu stackups, we have to add plies of high resin content prepregs to ensure there's enough resin to flow into the etched spaces between the copper patterns on the core layers during lamination and have enough resin (but- tercoat) between the glass and Cu to prevent failures and voiding (Figure 7). Holden: It can be different for each fabricator, even though they're using the same materials, because of process variations. Ellis: Exactly. In fact, it's one of the biggest complaints that Rich- ard and I hear for transfer stack- ups. TTM has a trademark called Seamless Global Transfer, which is one of our core competencies that we advertise. We can match stackups between our sites as closely as possible, so that we know that we do have the long- term prototype-to-volume pro- duction migration path planned and approved before we even start the first prototypes. But frustrated customers some- times ask, "Well, you're TTM. Why aren't all your stackups exactly the same?" We respond that they are not exactly Figure 6: A four-lamination cycle HDI stackup. Figure 7: Copper thicknesses and plies of prepregs.